ABSTRACT NUMBER: ESOC2026A1553 NORMAL INTERHEMISPHERIC CEREBRAL BLOOD FLOW ASYMMETRY - IMPLICATIONS FOR DEFINING THE NO-REFLOW PHENOMENON

Abstract Background and aims Definition of cerebral no-reflow commonly relies on a >15% cerebral blood flow (CBF) asymmetry threshold derived from historical HMPAO-SPECT studies. However, CT-perfusion (CTP) reference values are lacking. We investigated the normal range of physiological interhemispheric CBF variation in the human brain on CTP in patients without perturbed cerebral hemodynamics. Methods We retrospectively analysed an age-stratified random sample of adult patients presenting with acute neurological symptoms between January 2024 and August 2025 who underwent CTP with RAPID-processed CBF maps as part of a code-stroke protocol at a high-volume comprehensive stroke centre and were ultimately diagnosed with stroke mimics. Patients with significant internal carotid artery stenosis, intracranial lesions, prior stroke, cerebral amyloid angiopathy, or final diagnosis of migraine or seizure were excluded. A modified Harvard-Oxford atlas-based mask was applied to define eleven anatomical regions of interest (ROIs) per hemisphere. Median CBF values were extracted from each ROI, and interhemispheric asymmetry was quantified using right-to-left CBF ratios (Fig. 1). Results Among 200 patients (median age 72yo; 64.5% female), the median interhemispheric CBF ratio was 1.003, with pooled regional percentiles (2.5th/97.5th) indicating physiological asymmetry of ~13-15%. Variability was numerically greater in white matter than deep grey matter (~12-15% vs ~9-11%), although median absolute asymmetry did not differ significantly (p=0.76). Interhemispheric asymmetry varied across anatomical regions (p<0.001), being greatest in the insula (~20-21%) and lowest in the cerebellum (~8-10%). Conclusions Physiological variation in CBF between hemisphere on CTP is ~13-15%, supporting the >15% threshold to define cerebral no-reflow. Conflict of interest Maria Constanza Domine: nothing to disclose Figure 1 - belongs to Methods Figure 2 - belongs to Results